Inside the AI Startup School #2

The sessions were deliberately intimate and off the record – no recordings, no replays – just unfiltered advice in a room full of builders. But here are three high level insights we can share:

Don’t build broad – build sharp.

Speakers shared war stories of trying to be everything to everyone – and ending up with bloated, unfocused products that delighted no one. Brivael Le Pogam (CTO, Argil) described how their earliest prototype tried to handle multiple workflows, appeal to multiple personas, and show off every feature. But no one really understood what it was for.

Their turning point? Keep narrowing down to a single, specific use case – one that makes users say “finally.” The moment they did that, usage picked up, retention improved, and word-of-mouth started working.

In a world where attention is scarce and competitors are just a click away, focus isn’t a nice-to-have, it’s a survival strategy. Vagueness kills momentum, sharpness creates pull.

Pricing is positioning – and sometimes, cheap is worse.

Gabriel Hubert, cofounder & CEO of Dust started building an AI-powered assistant for CEOs and priced it at $10/month. It was technically impressive, worked well in tests… and no one bought it.

Why? Not because the product was bad – but because the pricing undercut its value. Execs saw the low price and assumed it couldn’t possibly solve an important problem. It felt like a toy. So the team rebranded, repositioned, and raised prices. Suddenly, the same product was seen as serious. Conversion rates went up.

Your price tells a story. Not just about affordability, but about ambition, value, and who your product is really for. 

Stop pitching. Start listening.

This theme came up again and again, especially in the early sessions: founders talk too much about what they’re building, and not enough about what users are feeling.

Jean-Louis Quéguiner (CEO, Gladia) talked about the turning point in their early user interviews: shifting from pitching features to asking blunt, open-ended questions. “What’s the most painful part of your current workflow?” “What did you try that didn’t work?” “What would break if this tool disappeared?”

These questions unearthed surprising insights – often unrelated to what the team initially thought was valuable. They stopped trying to “sell the product,” and started hunting for pain. That’s when real PMF signals began to emerge.

The takeaway: don’t ask people what they think of your idea. Ask them what they can’t live without. What they do today instead. What they duct-tape together. The gold is there.

The AI Startup School highlighted significant shifts in the AI Startup scene:

From “What’s possible?” to “What’s painful?”: successful founders prioritize solving specific user problems over flashy tech. They engage with users, ask tough questions, and develop solutions that are significantly better than existing options.

From “Go big fast” to “Earn the right to scale”: Instead of raising large funding rounds initially, many teams start small with trusted users. They handle customer support themselves and focus on usage metrics over vanity metrics.

From “AI as a Gimmick” to “AI as Infrastructure”: AI will power a wide range of applications, from robotics simulation tools to medical training assistants. Founders view AI as a powerful, invisible infrastructure, not the product itself.